DNA提取及通用引物选择对于PCR-DGGE对群落分析结果的影响

V ol. 25 No. 3, P. 310-316, 2007

DOI: 10.1007/s00343-007-0310-7

Effects of DNA extraction and universal primers on 16S rRNA gene-based DGGE analysis of a bacterial community from fish farming water*

LUO Peng (罗鹏), HU Chaoqun (胡超群)**, ZHANG Lüping (张吕平),

REN Chunhua (任春华), SHEN Qi (沈琪)

(The Key Laboratory of Applied Marine Biology of Guangdong Province and Chinese Academy of Sciences, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China)

Received May 10, 2006; revision accepted July 20, 2006

Abstract Among many reports investigating microbial diversity from environmental samples with denaturing gradient gel electrophoresis (DGGE), limited attention has been given to the effects of universal primers and DNA extraction on the outcome of DGGE analysis. In this study, these effects were tested with 16S rRNA gene-based DGGE on a bacterial community from farming water samples. The results indicate that the number of discernable bands in the DGGE fingerprint differed with the primer pairs used; the bands produced by 63f/518r, 341f/926r and 933f/1387r primer pairs were obviously fewer than those by 968f/1401r. Also, we found that each DNA extraction method resulted in different community profiles, reflected by the number and intensity of bands in the DGGE fingerprint. Furthermore, the main bands (theoretically representing dominant bacteria) differed with the extraction methods applied. It is therefore believed that the effects of universal primers and DNA extraction should be given more attention and carefully chosen before performing an investigation into a new environment with DGGE. Keyword: DNA extraction; universal primers; bacterial community; DGGE

1 INTRODUCTION

In the last 20 years, new advances in knowledge on microbial community profiles have been made with the application of molecular techniques such as denaturing gradient gel electrophoresis (DGGE), ribosomal intergenic spacer analysis (RISA), amplified ribosomal DNA restriction analysis (ARDRA), and single strand conformation polymorphism (SSCP). Among them, DGGE has become a popular method since it provides a rapid survey of microbial community by separating PCR (polymerase chain reaction) products according to melting behaviour of different amplicons from 16S rRNA gene (Muyzer and Smalla, 1998). However, among many reports investigating microbial diversity from environmental samples with PCR-DGGE, little attention has been paid to pitfalls in PCR-DGGE, which may potentially cause inaccurate biodiversity information. Some researchers have demonstrated artifacts caused by PCR amplification of mixed templates encompassing preferential amplification of certain sequences, chimeric amplification products, and erroneous nucleotides (Wintzingerode et al., 1997), but the effects of the method of DNA extraction and choice of universal primers on the outcome of indigenous microbial community analysis have not yet been discussed or established widely and in detail. Liesack et al. (1997) reported that environmental DNA extracted from a same soil sample using different lysis protocols produced different fingerprints with DGGE analysis. Lipthay et al. (2004) also found that each DNA extraction method resulted in a unique community profile from that of soil samples as measured by DGGE. In addition, RISA also found the bias caused by DNA extraction (Martin-Laurent et al., * This work was funded by the Special Scientific Foundation of Guangdong Province (No. A305030301) and was partly supported by a grant from KLFEE, Ministry of Agriculture (2003-04).

** Corresponding author: cqhu@scsio.ac.cnNo.3 LUO et al.: Effects of DNA extraction and universal primers on 16S rRNA gene-based DGGE analysis 311

2001). However, these researches concentrated on the effects of DNA extraction from soil samples. So far, no study has focused on the effects of DNA extraction from other environmental samples and on the selection of universal primers with DGGE analysis for microbial community profiles. It seems that little consideration has been taken for the influences of different DNA extraction techniques and selection of universal primers before PCR-DGGE operation in most literature on microbial ecology.

In this paper, the effect of different DNA extraction methods on DGGE patterns using environmental water samples was studied. The study demonstrated the effect of DNA extraction in the analysis of bacterial communities from soil DNA extraction. The effect of primer selection on the DGGE pattern of a bacterial community is also discussed.

2 MATERIALS AND METHODS

2.1 Environmental water sample

3000 ml of fresh water sample, collected from a grass fish (Ctenopharyngodon idella) pond in Guangzhou, China, was centrifuged at 10 000 g for 10 min (4°C). Pellets were resuspended in 50 ml of sterile water and divided into 40 subsamples. These subsamples were centrifuged again and the pellets (containing bacteria) were weighed and stored at −80ºC for subsequent DNA extraction.

2.2 DNA extraction from water samples Overall, seven methods of bacterial DNA extraction were used. Methods 1 and 2 are from commercial kits of bacterial DNA extraction, while Methods 3, 4, 5, 6 and 7 are from reported protocols of bacterial DNA isolation. Triplicate samples were used for every DNA extraction method.

Methods 1 and 2 were commercial procedures provided by the manufacturers of bacterial DNA extraction kits: the Huashun Mini Bacterial DNA Isolation Kit (Huashun Bioengineering, Shanghai, China) and the TaKaRa MiniBEST Bacterial Genomic DNA Extraction Kit (TaKaRa, Dalian, China). The two manufacturers did not provide complete information and the ingredients of their commercial kits. Method 3 is the modification of a previous protocol (Orisini et al., 2001). In brief, the pellet (prepared in section 2.1) was suspended in 300 μl of lysis solution and the suspension was microwaved for 45 s thrice at the highest power (2 500 W) at 5-min intervals. The suspension was then centrifuged at 12 000 g for 10 min, and 300 μl of phenol-chloroform (1:1, v/v) was added to the supernatant, put into a new tube, and then mixed thoroughly. The upper aqueous phase was piped for further extraction with phenol-chloroform. After the second extraction, DNA in aqueous phase was precipitated with absolute ethanol at 14 000 g for 10 min and washed twice with 70% ethanol. Finally, DNA was resolved in 50 μl TE buffer (pH 8.0). In Method 4, DNA was extracted according to a published protocol (Kalia et al., 1999). After the extraction, dried DNA was suspended in 50 μl TE buffer. In Method 5, DNA was extracted according to a modified phenol-chloroform-isoamyl alcohol method by Trochimchuk et al. (2003). The pellet of each sample was resuspended in 300 μl of GuSCN lysis buffer for 20 min at 65°C. After centrifugation, DNA in the supernatant was extracted twice with phenol-chloroform-isoamyl alcohol (25:24:1, in volume). Another operation was carried out in the light of the report (Trochimchuk et al., 2003). Method 6 was also a modified protocol of bacterial DNA extraction after Li et al. (2003). Bacterial sample in each pellet was suspended in 200 μl TE buffer and subjected to five freeze-thaw cycles, alternating between −80°C for 5 min and 65°C for 5 min. The suspension was centrifuged and lysis of unbroken cells was carried out according to Li et al. (2003). Both supernatants containing DNA were pooled for a subsequent procedure identical to that in Method 3. Method 7 was modified after Martin- Laurent et al. (2001). The pellet of each sample was suspended in 300 μl of lysis buffer for 20 min at 65°C, and the suspension was centrifuged at 12 000 g for 10 min. The supernatant was then treated as described in Method 3.

2.3 Measurement of DNA concentration and

purity

The concentration and purity of DNA isolated were determined spectrophotometrically (HITACHI, Japan) by taking the absorbance at 260 nm (A260) and 280 nm (A280).

2.4 Detection of genomic DNA obtained from different extraction methods

For each method tested, the presence and quality of extracted genomic DNA from two of the triplicate samples were analyzed on a 0.5% agarose gel containing 0.5 μg/ml ethidium bromide. 6 μl of each extracted DNA were added into the gel and electrophoresed for 30 min at 200 V followed byCHIN. J. OCEANOL. LIMNOL., 25(3), 2007 V ol.25 312

observation under UV light (VIBER LOURMAT, France). Unless otherwise mentioned, the following molecular procedures were carried out with extracted genomic DNA from two of the triplicate samples.

2.5 Selection of universal primers, PCR and DGGE

Four pairs of universal primers, anchoring different regions of 16S rRNA gene, were tested for their effect on the analysis of bacterial community profiles with PCR-DGGE. They were 63f/518r (Rombaut et al., 2001), 341f/926r (Massana et al., 2000), 933f/1387r (Araya et al., 2003) and 968f/1401r (Evans et al., 2004). A 40-nucleotide GC clamp was attached to the 5’-end of every forward primer to confer melting stability to the PCR products during DGGE. Compositions of the PCR mixtures were similar to the corresponding description in the references mentioned above, except that the concentration of DNA templates was adjusted to a final 1ng/μl. DNA from Method 1 was used as template for every PCR operation. PCR programs and electrophoresis of amplicons applied were in accordance with the corresponding original reports, in which above universal primer pairs were adopted.

Similarly, DGGE procedures were identical to those in the corresponding reports, in which the above universal primer pairs were adopted. PCR products were loaded to acrylamide gels in aliquots of 16 μl per lane and DGGE was performed with D-Code 2 000 system (Bio-Rad). After electrophoresis, the gel was stained with ethidium bromide and visualized under UV light (VIBER LOURMAT).

2.6 PCR-DGGE with 968f/1401r primers and DNA templates obtained from different extraction methods

According to the result from section 2.5, the 968f/1401r primer pair was selected to test the effect of different methods of DNA extraction on the DGGE fingerprint. To eliminate the influence of different concentrations of DNA template on PCR result, all the DNA templates from different extraction procedures were adjusted to 50 ng/μl in concentration prior to PCR. 1 μl of adjusted DNA template was used in each PCR mixture. To confirm the existence of amplifications, 3 μl of each product was added into a 1.0% agarose gel and electrophoresed for 20 min at 200 V followed by examination under UV light, for performing DGGE as per Evans et al. (2004). 2.7 Analysis of DGGE gels

Gel images were analyzed using LabImage Program 2.7.1 (Kapelan GmBH, Halle, Germany). Bands in DGGE fingerprints were automatically identified. Lanes were individually converted to filled plots by the program. After a background correction was made, the intensity of each band was measured by integrating the area under the peak and expressing the total area in the lane in percentage.

3 RESULTS

3.1 Yield, purity and quality of DNA obtained by different methods

Yield and purity of DNA obtained from seven extraction methods and the quality of genomic DNA are shown in Table 1 and Fig.1, respectively. Although high-molecular weight DNA (about 20 kb) could be acquired in 7 methods (Fig.1), different methods gave rise to yield and purity discrepancies. For instance, Method 4 yielded the most amount of DNA (mean 700.08 μg/g pellet), but concomitant with lower purity

of DNA (mean 1.75). Method 1 yielded the least amount of DNA (mean 120.52 μg/g pellet), but highest purity of DNA (mean 1.96).

Table 1 DNA yields and purity obtained by seven DNA extraction methods*

Method

DNA yield

(μg/g pellet)

DNA purity

(OD260/OD280)

1 120.52±11.73 1.96±0.08

2 286.95±31.67 1.62±0.05

3 590.05±79.68 1.87±0.02

4 700.08±37.00 1.75±0.06

5 168.42±12.99 1.86±0.04

6 210.08±17.9

7 1.68±0.07

7 326.31±26.19 1.82±0.03

*Data are represented in mean±standard error (n=3).

3.2 The effect of universal primer pairs on the DGGE pattern of a bacterial community

To assess the optimal primer pair used for DGGE analysis on bacterial community of the water sample, the 4 pairs of universal primers above were selected since they were frequently used in DGGE analysis of such communities. The result indicated that the number of bands detected by the LabImage software in DGGE fingerprinting differed with the primer pairs

No.3 LUO et al.: Effects of DNA extraction and universal primers on 16S rRNA gene-based DGGE analysis

313

Fig.1 Agarose gel electrophoresis of 6 μl DNA from each sample extracted using Methods 1-7

DNA from two of the triplicate samples (unequal weight) was loaded. Lanes: M, λ-Hind Ⅲ digest DNA marker (TaKaRa); 1-2, Method 1; 3-4, Method 2; 5-6, Method 3; 7-8, Method 4; 9-10 Method 5; 11-12, Method 6; 13-14, Method 7

used; 63f/518r, 341f/926r, 933f/1387r and 968f/1401r produced 10, 10, 21 and 28 discernable bands on average (Fig.2) respectively, with the reported DGGE procedures. 968f/1401r produced the most bands in DGGE with the same DNA sample. Therefore, the 968f/1401r primer pair was selected as PCR primers to assess the effect of 7 methods for DNA extraction on

the DGGE pattern of the bacterial community.

Fig.2 DGGE fingerprinting of PCR products obtained

using different universal primer pairs

Four primer pairs were tested in parallel for band numbers they gave rise to. DNA extracted from Method 1 was used in PCR

Lanes: 1-2, obtained using 63f/518r primers; 3-4, 341f/926r; 5-6, 933f/1387r; 7-8, 968f/1401r.

3.3 The effect of 7 DNA extraction methods on analysis results of the bacterial community profile

The effect of DNA extraction procedures on the result of a bacterial community profile analysis was studied through DGGE of 16S rRNA gene fragment amplified with 968f/1401r primers. Before DGGE, the existence of amplifications was confirmed by agarose gel electrophoresis. All extracted DNA from 7 methods could be effectively amplified (Fig.3). We found that each DNA extraction method displayed different community profiles. It was revealed not only by the number of DNA bands, but also by the intensity of DNA bands in DGGE gel (Fig.4). Among the 7 methods, Method 3 resulted in 30 bands (the most number) with DGGE on average, while Method 7 generated 16 bands (the least number); Methods 1, 2, 4, 5, and 6 produced 28, 21, 17, 17, and 19 bands, respectively. The difference in the number of bands detected was mainly caused by the occurrence or absence of some relatively weak bands. For example, in Method 1, 4 unique bands (A, B, C and D) could be identified at the nether part of Lane 1, but not in Method 5 (Figs.4 and 5). Bands in the same aclinic position (theoretically representing the same bacterial species) in DGGE gel had different relative intensities; abundance of a certain bacterium species in the bacterial community varied because of the difference

among DNA extraction methods. Even dominant

Fig.3 Agarose gel electrophoresis of PCR products with template DNA obtained by 7 extraction methods

968f/1401r primers were used. Lanes: M, DNA marker DL2000; 1-2, DNA extraction with Method 1; 3-4, Method 2; 5-6, Method 3; 7-8, Method 4; 9-10, Method 5; 11-12, Method 6; 13-14, Method 7

CHIN. J. OCEANOL. LIMNOL., 25(3), 2007 V ol.25

314

Fig.4 DGGE gel illustrated the different profiles of bacterial community caused by discrepance of DNA extraction

Lanes: 1-2, DNA extraction with Method 1; 3-4, Method 2; 5-6, Method 3; 7-8, Method 4; 9-10, Method 5; 11-12, Method 6; 13-14, Method 7

Fig.5 Comparison of filled plots of Lane 1, Lane 7,

and Lane 10 in DGGE image generated by

LabImage program

a. Filled plot of Lane 1;

b. Filled plot of Lane 7;

c. Filled plot of Lane 10

the uppermost section of gel, and its relative intensity in Lane 10 was 11.7%. In Methods 1, 4 and 7, the F band was the dominant band, and its relative intensity in Lane 7 was 11.1%.

4 DISCUSSION

Although numerous works have been done on

bacteria diversity in some environments by 16S rRNA gene-based DGGE analysis, little attention has been paid to the selection of universal primers in PCR-DGGE and the effect of DNA extraction on the analysis of a bacterial community. Seemingly, no single and simple reason can explain why a set of primers should be used in a particular study on bacterial community. Marchesi et al. (1998) reported that 63f/1387r primers (new designed primer pair) amplified the 16S rRNA gene from a wider range of bacteria than 27f/1392r primers, although the former showed some theoretical bias. Using different universal primers, Watanabe et al. (2001) found that some primer-specific bands appeared in the DGGE fingerprints. These findings demonstrate that empirical testing of universal primers is essential to confirm the applicability of PCR for different environmental samples.

In this study, we selected 4 pairs of universal primers that have been frequently used in PCR-DGGE analysis for samples from various environments. They showed different ability levels in amplification for the same DNA template extracted; among them, 968f/1401r gave rise to the most bands in DGGE analysis (DNA extracted in Method 1 was used as template). The discrepancy in the number of identified bands did not change with the concentration of DNA

templates used in PCR (data not shown). Therefore, we exclude the influence of concentration of DNA templates and the result may be explained using some hypothetical mechanisms. First, the differences in length, G+C content, sequence, and interaction of primers may influence the efficiency of amplification in PCR systems. This proposal is supported by Suzuki and Giovannoni (1996), whose comparison of the 27f/338r primer pair with 519f/1406r found that the latter showed lower amplification efficiencies. Second, these primers have been constructed theoretically using the (incomplete) database of 16S rRNA sequences from cultured organisms, and have not been tested systematically (Marchesi et al., 1998); these so-called universal primers may not actually be universal. The 16S rRNA gene sequences of some newly recognized groups are very diverse and include mismatches to the primers described above (von Wintingerode et al., 2000). As a consequence, for a specific primer pair, it possibly combines only part of 16S rRNA gene templates isolated from an environment of unknown biodiversity. Third, the 16S rRNA gene regions anchored by these primer pairs may have dissimilar extent of variation; the region anchored by 968f/1401r is likely quite variable compared with the regions anchored by other universal primers (Watanabe et al., 2001). In other words, although fewer bands were shown by PCR-DGGE with the other 3 pairs of primers, one band in the DGGE gels may represent several different 16S rRNA gene sequences that have similar melting points. Therefore, we believe that when a bacterial community from a new environment is investigated with PCR-DGGE, different primer selections should be considered for better prokaryotic taxon coverage of a specific environment, instead of doing only theoretical analysis and referring to related reports.

As a result, there is hardly a single method of DNA extraction applicable in analysis for all environmental samples with PCR-DGGE. The result of this study indicates that a bacterial community profile reflected by the DGGE band pattern is highly influenced by different DNA extraction methods. Different DNA extraction methods lead not only to the discrepant band number identified, but also the relative intensity of bands observed earlier (Lipthay et al., 2004). Therefore, it is suggested that the bias of DNA extraction from environmental samples is not occasional. In our experiment, except for the difference in DNA templates of the 7 methods of extraction, all the components of PCR mixture and PCR program were identical. Thus, causes are most likely within the extracted DNA samples themselves but not the PCR process. Alternatively, DNA samples would acquire different properties after DNA extraction with varying methods. The different properties probably lie in two aspects. First, DNA extraction may change the phylotype abundance and composition of the indigenous bacterial community (Martin-Laurent et al., 2001), which may be caused by varying degrees of lysis efficiency of different extraction methods or by different susceptibilities of diverse microorganisms in samples (More et al., 1994). In this study, the difference in the number of bands that could be detected was mainly due to the occurrence or absence of some relatively weak bands. Thus, less numerous bacterial types might have been differentially extracted using several methods. Second, the extraction may result in different levels of inhibitors in DNA samples. Inhibitors such as humic acids or heavy metals in natural samples have been reported and can apparently reduce the efficiency of PCR amplification. Therefore, these inhibitors with different levels in DNA samples might generate the discrepant efficiency of PCR amplification, which would be finally reflected in the variation of DGGE fingerprint.

In addition, Method 4 yielded the most extracted DNA but concomitant with the least DNA purity and very few DGGE bands. Method 3 yielded less extracted DNA than Method 1, but had higher DNA purity and resulted in the most number of bands in the DGGE gel. This indicated that a higher DNA yield does not always mean greater species richness (Stach et al., 2001). The yield and purity are equally important in conducting an objective assessment of a bacterial community with PCR-DGGE. Therefore, a method to produce the most DNA and highest DNA purity should be developed for DNA extraction from environment samples in the near future. In this study, combining Methods 3 and 4 may provide a good procedure for DNA extraction from fish farming water. However, even if a perfect DNA extraction method is developed for a specific environment, one cannot guarantee the accurate analysis of bacterial communities through 16S rRNA gene-based DGGE. In some species of bacteria, there are multiple copies of 16S rRNA genes with high heterogeneity (Dahllof et al., 2000; Mota et al., 2005). Thus, the number and abundance of “species” in community (revealed by DGGE fingerprint) may be inaccurately evaluated, which makes it difficult to interpret the 16S rRNA gene-based PCR-DGGE result (Dahllof et al., 2000). In a way, other genes such as gyrB (Tacâo et al., 2005) and rpoB (Dahllof et al., 2000; Mota et al., 2005) may

CHIN. J. OCEANOL. LIMNOL., 25(3), 2007 V ol.25 316

be an alternative for 16S rRNA gene in molecular analyses on bacterial communities.

The results of the present study indicated in general that the choice of universal primers and different methods of DNA extraction affected the DGGE pattern of a bacterial community in farming water. These effects, combined with artifacts by PCR amplification (Wintzingerode et al., 1997) and the aforementioned presence of multiple copies and heterogeneity of 16S rRNA genes (Dahllof et al., 2000; Mota et al., 2005), suggest that the 16S rRNA gene-based DGGE approach is of relative value when used for the analysis of environmental bacteria diversity. Thus, effects caused by universal primers and methods of DNA extraction should gain more attention. Primer pairs, which can produce the most number of bands in a DGGE gel, and a DNA extraction method, which can lead to the complete lysis of bacterial cells and highest DNA purity, should be explored before carrying out an investigation into a new environment through DGGE.

References

Araya, R., K. Tani, T. Takagi, N. Y amaguchi and M. Nasu, 2003.

Bacterial activity and community composition in stream water and biofilm from an urban river determined by fluorescent in situ hybridization and DGGE analysis. FEMS

Microbiol. Ecol.43: 111-119.

Dahllof, I., H. Baillie and S. Kjekkeberg, 2000. rpoB-based microbial community analysis avoids limitations inherent in 16S rRNA gene intraspecies heterogeneity. Appl.

Environ. Microbiol.66: 3376-3380.

Evans, F. F., A. S. Rosado, G. V. Sebastian, R. Casella, P.

Machado, C. Holmström, S. Jelleberg, J. D. van Elsas and L. Seldin, 2004. Impact of oil contamination and biostimulation on the diversity of indigenous bacterial communities in soil microcosms. FEMS Microbiol. Ecol.

48: 1-11.

Kalia, A., A. Rattan and P. Chopra, 1999. A method for extraction of high-quality and high-quantity genomic DNA generally applicable to pathogenic bacteria. Anal. Biochem.

275: 1-5.

Li M., J. H. Gong, M. Cottrill, H. Y u, C. Lange, J. Burton and E.

Topp, 2003. Evaluation of QIAamp® DNA stool mini kit for ecological studies of gut microbiota. J. Microbiol.

Methods54: 13-20.

Liesack, W., H. Weyland and E. Stackebrandt, 1991. Potential risks of gene amplification by PCR as determined by 16S rDNA analysis of a mixed-culture of strict basophilic bacteria. Microbiol. Ecol.21: 191–198.

Lipthay, J. R., C. Enzinger, K. Johnsen, J. Aamand and S. J.

Sørensen, 2004. Impact of DNA extraction method on bacterial community composition measured by denaturing gradient gel electrophoresis. Soil Biol. Biochem.36:

1 607-1 614. Marchesi, J. R., T. Sato, A. J. Weightman, T. A. Martin, J. Fry, S.

J. Hiom and W. G. Wade, 1998. Design and evaluation of

useful bacterium-specific PCR primers that amplify genes

coding for bacterial 16S rRNA. Appl. Environ. Microbiol.

: 795-799.

Martin-Laurent, F., L. Philippot, S. Hallet, R. Chaussod, J. C.

Germon and G. Soulas, 2001. DNA extraction from soils:

old bias from new microbial diversity analysis methods.

Appl. Environ. Microbiol.67: 2 354-2 359.

Massana, R. M. and C. Pedrós-Alió, 2000. Spatial differences in bacterioplankton composition along the Catalan coast (NW

Mediterranean) assessed by molecular fingerprinting.

FEMS Microbiol. Ecol.33: 51-59.

More, M. I., J. B. Herrick, M. C. Silva, W. C. Ghiorse and E. L.

Madsen, 1994. Quantitative cell lysis of indigenous microorganisms and rapid extraction of microbial DNA from sediment. Appl. Environ. Microbiol.60: 1 572-1 580. Mota, F. F., E. A. Gomes, E. Paiva and L. Seldin, 2005.

Assessment of the diversity of Paenibaccillus species in

environmental samples by a novel rpoB-based PCR-DGGE

method. FEMS Microbiol. Ecol.53: 317-328.

Muyzer, G.. and K. Smalla, 1998. Application of denaturing gradient gel electrophoresis (DGGE) in microbial ecology.

Anton. Leeuw. Int. J. Gen. Mol. Methods56: 365-373. Orisini, M. and V. Romano-Spica, 2001. A microwave-based method for nucleic acid isolation from environmental samples. Lett. Appl. Microbiol.33: 17-20.

Rombaut, G.., G. Suantika, N. Boon, S. Maertens, P. Dhert, E.

Top, P. Sorgeloos and W. V erstraete, 2001. Monitoring of

the evolving diversity of the microbial community present

in rotifer cultures. Aquaculture198: 237–252.

Stach, E. M. J., S. Bathe, J.P. Clapp and R.G. Burns, 2001.

PCR-SSCP comparison of 16S rDNA sequence diversity in

soil DNA obtained using different isolation and purification

methods. FEMS Microbiol. Ecol.36: 139-151.

Suzuki, M. T. and S. J. Giovannoni, 1996. Bias caused by template annealing in the amplification of mixtures of 16S

rRNA genes by PCR. Appl. Environ. Microbiol.62:

625-630.

Tacâo, M., A. Moura, A. Alves, I. Henriques, M. J. Saavedra and

A. Correia, 2005. Evaluation of 16S rDNA- and

gyrB-DGGE for typing members of the genus Aeromonas.

FEMS Microbiol. Lett.246: 11-18.

Trochimchuk, T., J. Fotheringham, E. Topp, H. Schraft and K.T.

Leung, 2003. Comparison of DNA extraction and purification methods to detect Escherichia coli O157:H7 in

cattle manure. J. Microbiol. Methods54: 165-175.

von Wintzingerode, F., O. Landt, A. Ehrlich and U. B. GÓbel, 2000. Peptide nucleic acid-mediated PCR clamping as a

useful supplement in the determination of microbial diversity. Appl. Environ. Microbiol. 66: 549-557. Watanabe, K., Y. Kodama and S. Harayama, 2001. Design and evaluation of PCR primers to amplify bacterial 16S ribosomal DNA fragments used for community fingerprinting. J. Microbiol. Methods44: 253-262.